How Do Metagames Work And How Can You Solve Them?

You will have heard a lot talk about the metagame if you follow Constructed strategy articles – especially lately in the context of Regionals. In fact, the metagame is one of the favorite topics of discussion for Magic theorists, tournament players, and people who want to sound like they are in the know… But what exactly is it? Not many people can actually give a satisfactory definition of the word, and will just mutter something about tier 1 decks and leave it at that. Instead, why not read this helpful introduction?

What Is The Metagame?

“Slide is a good metagame choice in an R/G-heavy environment.”

“The abundance of Compost in sideboards caused MBC to drop out of the metagame.”

“Tog is adaptable to be competitive in a variety of different metagames.”

“Odyssey Block Constructed was dull due to a stagnant metagame.”

You will have heard a lot of statements like this if you follow Constructed strategy articles – especially lately in the context of Regionals. In fact, the metagame is one of the favorite topics of discussion for Magic theorists, tournament players, and people who want to sound like they are in the know… But what exactly is it? Not many people can actually give a satisfactory definition of the word, and will just mutter something about tier 1 decks and leave it at that.

However, given that it is difficult to discuss something if we are not sure what exactly it is, I have turned it over to our friends at dictionary.com. After sifting through about twenty different translations, I came up with the following.

Meta – Beyond; transcending; more comprehensive

Game – A competitive activity or sport in which players contend with each other according to a set of rules

So the literal translation of metagame is”the game beyond.” This sounds a bit like a crap 70’s sci-fi series and also doesn’t mean an awful lot on its own, so let’s put it in context.

The metagame is the game that exists in Magic beyond Magic itself. This game is both related to, yet independent, from the rules of Magic. It is the game you play not at the game table but in your head, sometimes weeks before the tournament, not against one opponent but against all of them. It is the game of deck selection and deckbuilding.

When we are asking,”How do metagames work?” what we are really asking is,”What determines what decks are played at major tournaments?” It is this question that I will attempt to answer in this article.

If, when going to a Constructed tournament, you just throw sixty cards together that look cool or pull off an amusing combo then that’s fine – but I don’t think the rest of this article will be of much interest to you. If, however, you take a deck that you feel has the best possible chance against the decks that you are likely to face, then you are already playing the metagame and may want to read on.

A Simple Metagame

Metagames are normally very complex, but there are some exceptions. Odyssey Block Constructed, especially during the early part of the season, was a format where two decks dominated: Namely, U/G Madness and Mono-Black Control. There were various different builds of both, but these variations were not really different enough to be considered completely different archetypes. Various arguments were presented, arguing that one deck was more powerful than the other, but none were particularly convincing. Although more people on the whole played U/G, this was probably more to do with the cost of building the deck and the ease of playing than any particular power difference. In fact, for all intents and purposes, most people will admit that U/G had approximately a 50% win rate against MBC.

For most of the season nobody could find a deck that could go more than 50% against either of these decks without being seriously bad against the other. In this particular setup, there is no point in playing any deck other than these two – and because the decks are 50/50 in matches between the two, it matters little which deck you play. What we have here is a completely stagnant metagame – there are only two decks, and it doesn’t matter which you play as both win 50% of games against either deck.

Of course, this is a bit simplistic. If this was truly the case, then this would have been the most boring tournament scene ever, with play-skill and luck alone deciding who won the tournament. In reality, the discovery of new tech for both decks meant that deckbuilding was still important – although mainly this consisted of tweaking U/G or MBC to the best configuration possible. As an aside, what eventually happened was that three new decks – a U/W deck called”Punisher,” a control deck called”Wake Combo,” and a mono-black aggro deck called”Pirates” were developed, which spiced up the metagame for a while. Then Onslaught was released and time was called on a very boring season just as it was getting interesting.

A More Complex Metagame

This is a particularly extreme example, however. Normally, metagames are dynamic and involve a lot of complex decision making as people look to”solve the metagame.” Let’s take a look at the current Standard environment, for example…




Actually, let’s not. I’ve just looked at the current Standard metagame and it’s very messy. Instead, I have decided to cheat and make up a metagame of my own. This is a fantasy environment that, on the surface, is pretty close to the real one but is actually very simplified. Because it’s my own fantasy metagame, I am going to introduce some unrealistic assumptions to make the illustrations easier. These are:

  • There is only one build of each deck

  • There are no rogue decks

  • All players have the same level of play skill and all decks are equally easy to play

I know this is unrealistic, but it is necessary to keep this simply. In our made-up environment, there are going to be four decks:

  • U/G Madness (U/G)

  • R/G Beats (R/G)

  • Psychatog (Tog)

  • Astroglide (Slide)

These decks have the following win percentages against each other:

  • Slide wins against R/G 70% of the time

  • R/G wins against Tog 70% of the time

  • Tog beats Slide 70% of the time

  • U/G wins 50% of all games against the above 3 decks

  • Obviously in the mirror, all decks are 50%

So what is the best deck to play in this environment? Difficult to say. If you expect a lot of Tog, then R/G is the best choice. If R/G is big, then Slide is best. Let’s assume that you expect all decks to be played equally. The probability of winning any given match with U/G is obviously 50%. For the other three decks, your win percentage is as follows.

(0.25 * 0.7) + (0.25 * 0.3) + (0.25 * 0.5) + (0.25 * 0.5)

= 0.25 * (0.7 + 0.3 + 0.5 + 0.5)

= 0.25 * 2

= 0.5

So every deck will have, on average, a 50% chance of winning each match.

Now, the mathematicians amongst you (and most of the non-mathematicians to be honest) are probably hitting themselves over the head with a spoon right now. Yes, this is pretty obvious due to the symmetry of the model – but its understanding is vital before we go any further.

Every deck except for U/G has a good matchup, two average matchups (U/G and the mirror) and a bad matchup. In this situation, there is no tendency for any deck to be played more than any other – and just like in the simple metagame above, no deck choice is better than any other. However, as we will see, one minor change can make this metagame far more interesting.

Introducing Instability

Sticking with the model above, let’s say that Billy Pro-Tour comes up with a great new bit of tech for Psychatog so that its record against Slide is now 90% instead of 70%, and publishes it on StarCityGames so that everyone knows about it. If everyone still plays the decks in equal numbers, the win percentage for Tog is now:

(0.25 * 0.9) + (0.25 * 0.3) + (0.25 * 0.5) + (0.25 * 0.5)

= 0.55 or 55%

Meanwhile, the win percentage for Slide will fall to

(0.25 * 0.1) + (0.25 * 0.7) + (0.25 * 0.5) + (0.25 * 0.5)

= 0.45 or 45%

This means that all decks no longer have the same expected win percentage. Tog will win 55% of its matches while Slide falls to 45% and the other two decks remain at 50%. Therefore, Tog becomes the best deck in the environment and is the best deck to play at the tournament, right?

Well not quite. The problem is that because everyone is privy to this new bit of tech, then it is only fair to assume that most people will come to the same conclusion you have – namely, that Tog is the best deck in the format. Therefore, the assumption that all decks will be played equally is now unrealistic. Let’s say that 70% of players now play Tog, with the remaining 30% split between the other three decks. Without typing out all the math (I’m sure you’re getting bored with it by now), the expected win percentage for each deck is:

  • Tog: 52%

  • U/G: 50%

  • Slide: 24%

  • R/G: 62%

What’s happened here? We have made a well-publicized improvement to Tog and as a result have made R/G the best deck in the format. Of course, it doesn’t stop here. Smart players will realize this in advance and will play R/G. If there are enough of these smart players in the tournament, then maybe slide will become the best deck as a result.

Here we have the foundations of a dynamic metagame. When there is a situation like this, it is impossible to define the”best deck,” as what the best deck is depends entirely on the makeup of the field. When I first started playing Magic, one of the first things I read was this:

“At any major tournament, to do well you will need a deck that beats the best decks in the format. To actually win the tournament, you will need a deck that beats the decks that beat the best decks in the format.”

At the time I thought this was a bit of a meaningless soundbite, but using the example above, the truth of this statement shines.

More Metagame Complexity

Of course, we have still just only scratched the surface of the complexity of a real life metagame. There are may other factors which make the issue of”best deck” even more polarized. Let’s discuss some of them in a bit of detail.

Different Builds:

Obviously not all decks that are of the same archetype are exactly the same and this can affect certain matchups. For example, in the Psychatog mirror, if only one deck is running Cunning Wish then that deck will win more often than not. In U/G, Merfolk Looters are good in the mirror match but not so good against R/G. There are endless examples like this.

The important thing is that if a change to a deck improves some matchups without making others worse, then it is a unilaterally good change and should be included in any build of the deck (and therefore becomes the new standard). However, it is those that make one matchup better and another matchup worse that complicate the metagame. If, for example, some players replace the Merfolk Looters in their U/G deck with Aquamoebas, this means there are two different types of U/G deck with different win percentages against the field. This of course makes our generic ‘Deck A beats Deck B n% of the time’ statements pretty unreliable.

Rogue Decks:

Even at the highest level, some players will decide to”go rogue.” Sometimes this is because a player thinks they have discovered the next big thing, other times they just want to catch people of guard with something they won’t be prepared for. Until now we haven’t really considered the place of rogue decks in the metagame. Because by definition rogue decks are unknowns (or at least not highly regarded), it is difficult to assess how your deck will perform against them. What is useful, though, is to consider the robustness of your deck.

U/G, for example, is very robust. It plays very efficient threats very quickly and then counters anything that interferes with its plans. It can break ground stalls by flying over with Wonder and can kill problem enchantments with Ray of Revelation. It is reasonable for a U/G player to feel confident when playing against a rogue deck (with a few exceptions).

Now consider Tog. While Tog is good enough that it will destroy bad decks, it can have trouble beating weenie rushes that can play more than one threat per turn because Tog relies on trading one-for-one until it builds up the critical mana for the kill. Sligh, WW, and Elves.dec all give Tog headaches. Because they have bad matchups with U/G, R/G, and Slide these decks are not really in the metagame at the moment – but it is worth considering that Tog is probably less robust than U/G, a measure which doesn’t show up when you are looking at specific matchups. Nevertheless, it is significant.

Uncertainty Principle:

This is probably the overriding factor that makes any given metagame so difficult to model. Our example above has relied on us knowing which decks beat which other decks, and what percentage of the time. The fact is, you just can’t calculate this. Even simple questions such as,”Does Tog beat U/G?” have failed to produce a definitive answer. One testing group will find one result and then another will find a conflicting result. The reason for this is that there are too many variables in the game of Magic. Testing groups may be playing slightly different builds. Some players are simply better at playing certain decks, and so their results are skewed in favor of that deck. Sometimes you just suffer from statistical variation. Even if you play a matchup fifty times, if one deck wins thirty of those, it may either have a strong advantage or just have got the rub of the green in a close matchup. Because of this, the win percentage figures you use are only ever going to be best guesses – and as a result, conclusions you derive should be treated with caution, especially if the results are close.

Solving The Metagame

So far we have defined what a metagame is, why they occur and how they work. However, this is not terribly useful unless you can in some way use this information to your advantage. By now, you should have figured that this is a two-step process: Firstly, working out what the expected field is, and secondly, working out what deck to play to best exploit this knowledge. Let’s tackle the first of these in a bit more detail.

Let’s say that you are preparing for an important Constructed tournament and want to give yourself the best chance of winning. What resources should you be using to work out what the rest of the field will be playing? Well, here are some possibilities:

Look at what did well at the most recent Grand Prixs and Pro Tours. These are the tournaments that people pay most heed to when deciding what to play and it is these decks that define the metagame as most people know it.

Check out any deck lists posted on Sideboard or Star City. Often, people will copy these decks, especially if they receive an endorsement from a well-known player. [author name="Rob Dougherty"]Rob Dougherty’s[/author] Elvish Succession deck is an example here and has become a significant deck in the metagame in the month since it was posted. (Sadly, that is debatable – The Ferrett)

Scout your local tournaments. Obviously this is very useful if the tournament you are preparing for is a local tournament, as a lot of players play the same decks week after week. Even if it isn’t, you may get some idea of tendencies towards or away from certain decks.

Check out Magic Online. For those who use Magic Online, go to the Premier Events room and you can watch replays of the Top 8 matches from all the recent Constructed tournament played. For those who don’t use Magic Online and have no intention of ever using it, consider this: How would you like to pay $10 and get to watch the top 8 matches of around twenty constructed tournaments per week for ever from the convenience of your PC? Sound good? You will even get a $10 voucher thrown in for free so you can buy some cards to play around if you get bored.

Take specific metagame tendencies into account. Here is a quote from the master himself Kai Budde, from a sideboard article a while back:

“One of the old rules at European Magic events is that Scandinavians love their islands. Whenever you ran into one, they would be playing some sort of blue deck. With that in mind, I built a Psychatog version that was tuned to win the mirror match and smash anything like Wake and black control.”

Kai noticed that in any given metagame, Scandinavian players would be more likely to play the more control-oriented decks. Any similar observations you can make about the metagame you will be playing in are very valuable. For example, some have noticed for example that States and Regionals tend to be very beatdown-oriented. Others, that UK tournaments contain strong rogue decks. Any information like this can be useful.

This is not an extensive list, of course. Maybe you have other ideas. What you are hoping for is an idea of the field composition. You want to have ballpark figures for what percentage of the field should be playing what deck. This is important for the next stage.

So once you think you know what the field will look like, it’s time to build your deck. What you are looking for is a deck that has the highest possible win percentage against the field you expect to face. You have two options: You can either play the existing deck that you feel has the best chance against the field, or you can try and build a rogue deck that”breaks the metagame.”

Most people dream of doing the latter, but format-breaking rogue decks are hard to come by. There are full-time Magic players who have nothing to do except sit at home and try to find these decks. They are better players than most of us, and many of them go their entire careers without discovering a deck like this (one famous exception is Pro Tour: Tokyo, where Zvi Mowshowitz won with”The Solution”). The idea of winning with an established deck is far more realistic.

Obviously, on the minus side, people will be prepared for your deck and may have tech against it – and often you may have to rely on tight play and good matchups to go all the way. However, on the plus side, if you have predicted the metagame correctly, you are playing with a deck that is tried and tested and will have favorable matchups against the field in general. If you look at the tournament reports from big tournaments, you will see they are nearly all won by well-known deck archetypes that fared well against the rest of the field.

So how do we go about choosing one? Well first of all, for all decks you expect to be played in significant quantity (let’s say 5% of the field or more) you need the important information on which decks beat which other decks what percent of the time. Ideally, you would get this information by testing each matchup many times – but realistically, not all players have the time to do this (or just can’t be bothered) and so these figures may have to be gained through intuition or taken from a less-thorough testing gauntlet. Bear in mind, though, that the better your testing, the better your figures will be. Now make a table containing this information with each deck written along the top and down the side (Excel is good for this). Now for each row (deck) in the table, go along the columns (matchups) and write down what win percentage you expect for that matchup. When you have done this for each deck, you have a matrix of values broadly representing the metagame. For example, your table may read something like this:





































Now, all you need to do is multiply each figure with the fraction of people you expect to be playing that deck and then add all the figures together. For example, let’s say we expect the following field composition.

  • U/G: 30%

  • Tog: 15%

  • R/G: 15%

  • Slide: 10%

  • Wake: 20%

Note that these figures don’t add up to 100%. The remaining 10% in this example is made up of all the other decks in the field (Tier 2, rogue, and scrubby decks) that won’t make up a significant proportion of the field individually.

This is the exact same calculation we did for our example metagame before except now obviously we are using your real data. Note that the number you get will not represent your actual win percentage because the composition percentages do not add up to 100%. If this really bothers you, add”rogue” as an additional deck type and give it an arbitrary win percentage versus the whole field. In this example, I will pretend these five decks beat everything rogue 100% of the time. Note that it doesn’t matter that this is unrealistic, providing you use the same value for all decks. If you like, you can adjust the figures according to how robust you perceive the deck to be (as described above)…. But given that we are only talking about small percentages, it is probably not necessary and I won’t bother in this calculation. These are the calculations for the win percentages of each deck in this example.

U/G: (0.5 * 0.3) + (0.6 * 0.15) + (0.55 * 0.15) + (0.55 * 0.1) + (0.4 * 0.2) + (1 * 0.1) = 0.15 + 0.09 + 0.0825 + 0.055 + 0.08 + 0.1 = 0.5575

Tog: (0.4 * 0.3) + (0.5 * 0.15) + (0.3 * 0.15) + (0.9 * 0.1) + (0.7 * 0.2) + (1 * 0.1) = 0.12 + 0.075 + 0.045 + 0.09 + 0.14 + 0.1 = 0.57

RG: (0.45 * 0.3) + (0.7 * 0.15) + (0.5 * 0.15) + (0.3 * 0.1) + (0.35 * 0.2) + (1 * 0.1) = 0.135 + 0.105 + 0.075 + 0.03 + 0.07 + 0.1 = 0.515

Slide: (0.45 * 0.3) + (0.1 * 0.15) + (0.7 * 0.15) + (0.5 * 0.1) + (0.4 * 0.2) + (1 * 0.1) = 0.135 + 0.015 + 0.105 + 0.05 + 0.08 + 0.1 = 0.485

Wake: (0.6 * 0.3) + (0.3 * 0.15) + (0.65 * 0.15) + (0.6 * 0.1) + (0.5 * 0.2) + (1 * 0.1) = 0.18 + 0.045 + 0.0975 + 0.06 + 0.1 + 0.1 = 0.5825

Therefore, in this metagame (which is made up remember) the deck with the highest win percentage (in order) would be:

  • Wake: 58.25%

  • Tog: 57%

  • U/G: 55.75%

  • R/G: 51.5%

  • Slide: 48.5%

Therefore, if you want the best win percentage (in this particular made up metagame) then Wake would be the deck to go for. You may be excited enough about this conclusion that this will decide your deck there and then. However more likely you can use this conclusion along with additional factors that affect your choice. Specifically:

Some people simply prefer some deck styles and are better with them. You shouldn’t play a deck you are not comfortable with, even if it is statistically the”best deck.” You will likely play better and make fewer mistakes with a deck you are comfortable with and enjoy playing.

Some decks are more difficult to play than others. Always consider your own limitations and strengths when deciding your deck. For example, if you don’t know Wake inside-out, I would advise against playing it in a tournament.

Sometimes you may want to take a risk. Let’s say there is a deck that has one very poor matchup, but is otherwise very good. As a result, this deck comes up as average in the above analysis – but if it were not for the bad matchup, it would be the best. You may want to take this deck and risk that you will not come across your bad matchup and it may pay off. Of course, you may face it twice in the first two rounds and be faced with an 0-2 drop… But that’s why it’s called taking a risk.

By now, you should have a good idea of what deck you want to play and you have given yourself the best chance by picking the best deck for the expected metagame. Of course, you may want to tweak the deck to improve certain matchups or maybe even have a go at going rogue… But these are separate articles in themselves. Which deck to play is just too big a topic to analyze all in one go.

Hopefully, this article has given you ideas as to how you can make better deck selection choices. Maybe you think that this is just too imprecise a science with too many unknowns – and as a result, any conclusions are worthless; I can certainly sympathize with this argument. However, if nothing else it should have opened your mind to the possibilities of deck selection within a tournament environment – to the strange and mysterious dance that is the metagame.


Piemaster” on MTGO

[email protected]